EBT Motorcar M-3 Restoration
This page details the restoration work performed on the EBT railcar M-3 during 2005 and 2006. The goal of the work was to return the unit to safe and reliable service as a historic relic.
History of the M-3
From late 1923 to 1924 the East Broad Top constructed a small enclosed motorized unit for use in inspections of the line. Built primarily from Maxwell automobile parts and components fabricated in the EBT shops, the car was larger than the other maintenance cars the line used.
In 1928 the car was significantly rebuilt using the drive train and other parts from a Nash automobile. At that time (two years after the M-1 was built) the car was christened the M-3.
The M-3 served as an inspection vehicle during the rest of the railroad's operating lifetime. In 1960 when the EBT reopened as a tourist carrier, the M-3 was also re-operated. However the car was stored inoperative sometime in the 1960's. There were several stories about why. One was that it was run without enough oil and the engine seized. Another is that a replacement cam shaft was installed without being hardened enough and was ground down.
In 2005, with the approval of EBT management, the car was removed from the railroad for the first time in it's existence so that it could be restored to operation.
Once the M-3 was moved to a location where it could be worked on, the first order of business was to remove the engine and send it to a professional rebuilder, Rhinehold Restorations. At first the rebuilders had no idea what it was. After a thorough cleaning of the engine and inspection, the serial numbers showed that it was Nash, 1924 production. As the car seems to have received the 1924 engine in 1928, it would have been a used engine at the time it was installed. According to stories the donor car belonged to a railroad executive at the EBT.
The restoration shop disassembled the engine to get a better idea of what would be needed to return it to service. An initial inspection showed that neither of the rumored reasons the car was not operable were true. The engine turned fairly easily once the head was removed. Some valves were stuck to the head from lack of use which prevented the engine from turning until the head was removed. The cam shaft lobes were normal.
After the initial disassembly and cleaning, Heb Kephart, longtime Rockhill Trolley Museum volunteer, volunteered his extensive knowledge to implement the actual engine rebuild. The engine was transferred from Rhinehold Restorations to Herb's own shop where be set about the task of bringing the old Nash power plant back to life.
There were a number of issues, though, most involving the head and valves. At some time in the life of the engine the head dropped a valve into the #3 cylinder. When the piston came up it forced the valve into the head. To repair the head a slanted hole was bored through the head and a solid copper rod was inserted. This "injury" may have been the reason the engine was rendered surplus from its role in an automobile.
Unlike more modern engines, this one has no center bearing halfway down the crankshaft, leaving only two main bearings, one at each end. The main bearings were not worn out, but had been worn to a cone shape from flexing of the crankshaft. The flexing was a result of the engine being overreved, possibly due to high engine speed during backup moves (evading a steam train perhaps?) The backup gear has a high gear reduction. The crankshaft itself was of unusually heavy construction, with twice the mass of a Model T crankshaft.
It was found that the points in the distributor were the wrong ones for the engine, resulting in a spark that was not timed for the valve movement. This may be the real reason the car no longer ran after the 1960's.
During the rebuild, all the pistons, springs, vales, pistons and bearings were replaced. The original main bearings were insert, but replacements were poured. Mechanical replacement parts were used where they fit the engine. Toyota pistons, Cadillac valves and Chevrolet valve springs were close matches to the original Nash parts. All new spring tops were made and heat treated.
The original engine in the car was likely about a 1918 Maxwell. The 1918 engine had the carburetor on the right side of the engine. This is the side of the car that the fuel tank was mounted on. As these engines relied on gravity rather than a fuel pump to deliver gasoline to the carburetor, it would have been logical to have the tank as close to the carburetor as possible. The Nash engine, by contrast, has the carburetor on the left side, opposite of the tank. The decision was made during the restoration to relocate the tank to the left side to reduce the distance the fuel has to travel to the carburetor.
The fuel tank itself was another quandary. The original was made of riveted sheet metal. Adequate in its day, it posed some safety concerns being in the passenger compartment as well as potential leaks. For safety, the tank was replaced by a new one of the same dimensions constructed of all welded seams. Of additional concern was the means of fueling, which had been done by flipping up one of the passenger seats to access the top of the tank. Again for safety, it was elected to change this feature, adding a hidden external filler tube in the front wall. A modern safety cap, pressure release, and overflow were added to the tank to protect from any fueling accidents. The new fill hole was covered with a door to retain the car's original appearance.
Many parts for the engine were virtually impossible to find. The fan was only a two blade design and powered by single oversize V-belt. No automotive or machine shops had a similar belt, but one was located at an agricultural implements supply shop.
The radiator was recored and remained original, though with a slightly different core design. The shroud over the radiator was originally nickel plated. The nickel planting had long ago flaked off allowing rust to set in on the shroud. The shroud was cleaned down to bare metal and was coated with copper first then chromed (this is standard practice in chroming metal parts as chrome does not adhere well to bare metal.) Although more silver than the original nickel's slightly gold color, the chrome finish will last indefinitely, while nickel will tarnish and eventually flake.
The Nash emblem that had adorned the radiator shroud was long missing. An exact duplicate was found on eBay and installed on the shroud. A new radiator cap was machined from a block of aluminum to replace the missing original.
The coolant tubes mounted on the engine, to which the coolant hoses attach, were poorly aligned to match the actual direction the hoses traveled to the radiator. Both tubes were removed and angles cut and brazed to create the correct direction and minimize stress on the coolant hoses.
The distributor cap was missing from the car when the restoration began. 1924 Nash distributor caps are hard to come by. The only one that could be located had a section broken out of it. Carefully, the missing section was filled in with putty and sanded to shape to match the original shape.
With the recreation of the pilot on the car (more about that later) it was decided to also recreate the crank starter that resided in the pilot. The crank was built to look like the original in the builder's photo, and to be functional as well. The crank engages the flywheel on the front of the engine and is spring loaded to retract from the flywheel once the engine starts. Once the car was mostly assembled it was found the crank could start the engine in as little as a quarter of a turn.
Once running it was found that the engine ran very smoothly with little vibration. Initial plans to retrofit the car with more robust motor mounts were discarded when it was found how little vibration the engine exhibited when running. However, the original angle iron that supported the engine in the frame had become cracked at the bends. These cracks were repaired by welding.
The transmission in the car was the original Nash three speed manual that would have been mated to the engine during automotive service. Unlike modern manual transmissions, this one does not have automatic sychronization of the gears (a system to get the gears spinning the same speed before being meshed.) This requires the transmission to be "Double Clutched", or the cluth engaged both before leaving one gear and again before entering another.
An inspection of the transmission showed less than 10% tooth wear, which meant no gearing work was needed. New lining was installed on the clutch plate.
Currently the M-3 has a Ford Model A rear end. The unit previously seems to have had a Nash rear end. This is evident because when the Ford rear end was installed the railroad welded the spline from the Nash rear end to the Ford rear end. The change out may have been motivated by the fact that the Ford has lower gear reduction than the Nash, allowing higher speeds. The Nash rear end was presumably installed in 1928 with the Nash engine, transmission and drive shaft. It is not known when the current Model A rear end was installed. Operationally, the M-3 could travel about 24 MPH in third gear and 8 MPH in reverse at 1800 RPM.
Of the two old Buda wheels that were on the rear of the car, the right rear wheel was far more badly worn. The rear end of the car still had differential action, needed when an auto turns a corner so the outer wheel can follow a longer arc. Differential action also has the undesirable effect of causing only the right wheel to spin when traction is lost. This accounts for the much greater wear on the right rear wheel. In automobiles, this can be remedied with "limited slip" where a brake-like appliance keeps the wheels from turning at a vastly different rate. However, differential action is not needed on railroad vehicles. The curve arcs are far wider making the differential in wheel travel almost inconsequential. Also the slight taper on railroad wheels compensates for what differential there is. Therefore it was decided to "lock" the differential, causing both wheels to turn together at all times. This "Live Axle" keeps equal power applied to each of the two driving wheels. This will help in both acceleration and braking.
The main problem with the rear end was that the seals were a custom size rather than using off-the-shelf components. This made future maintenance problematic at best. In order to correct this problem, the tubes on either side of the differential were removed and tubes from a Ford F-250 rear end were installed in their place, providing a modern and robust structure that can better handle the load and track irregularities. Also when seals need to be replaced in the future, stock items can be used. The axles from the F-250 rear end were also installed into the differential. These axles had the added advantage of directly fitting the replacement Fairmont M-O-W wheels rather than requiring a weld job as the Nash axles did.
Suspension, Wheels and Brakes
Inspection of the car revealed many suspension problems. In addition to several broken leaves, the springs had lost some of their "spring" causing the car to ride low and the wheel flanges to eat away at the fenders. The front and rear spring mounts also had enough lateral slop that springs rubbed on and cut into the frame. This was due to undersize bolts installed in the spring linkage. The front axle had about one inch play side to side.
All four spring sets were replaced with new, modern leaf springs that cosmetically matched the originals. The leaves and bolts in all the spring rigging were replaced with those of the correct size.
Of the four wheels the two Fairmont wheels, on the front, were still in fairly good condition and were reportedly replaced in the 1960's (stamped June 1960). The other two which may have been Buda wheels, were badly worn and had been welded in their treads. They had also been welded in various questionable ways to the Nash axle hubs rather than bolted. The Buda style wheels were originally on all four hubs, but presumably the front two were replaced due to the same type of deterioration.
Through the generosity of volunteers and donors, two new Fairmont wheels were purchased at no small cost. The were used to replace the badly deteriorated Buda wheels. This gave the car a matched set of four low (or no) wear wheels with no welded repairs.
The car as built in 1924 had no brakes on the wheels as is traditional for similar railroad cars. Presumably the parking brake, which consists of a band brake on the back end of the transmission, was used exclusively. This must have proved inadequate, as during the 1928 rebuild six wheel brakes were added, two on each front wheel and one on each rear. The wheel brakes were manual, activated by a tall lever inside the car.
The old brake lining was found to be worn out and would do little to help stop the car. It appeared to be made from the same conveyor belt material that was used on the interior floor. The brake shoes were separating from the brake lining. New brake shoes and hangars were fabricated from oak and steel, and new modern brake lining was applied. Numerous new sleeves were installed throughout the brake linkages.
The electrical harness in the car was entirely replaced with modern wiring with more reliable insulation. The M-3 is unlike modern automobiles in that it has a 6 volt DC electrical system rather than 12 volt. Where possible original dashboard components were reused, but most electrical switches had to be replaced with new ones. Originally a key switch was used for the ignition switch with a rotating outside ring that activated the lights. This switch could not be repaired so was replaced with a hand switch that had one position for running and one for running with lights.
The headlight is an unusual design, where an adjustment extended through the front wall that allowed the position of the lamp to be faired, changing the tightness of the headlight beam. The headlight was disassembled and restored.
The M-3 had two horns through most of its life. An "Aooga" horn is mounted above the headlight in the front wall and a bellow horn was later mounted on the front left fender. The original Aooga horn was still on the vehicle when restoration began, though it was not functional. It was removed from the car and disassembled. Worn parts inside the horn were repaired or replaced and the horn was returned to service. When tested it sounded strangely like the "dive" horns on WWII submarines!
The fender horn was missing from the unit for many years and it's whereabouts unknown. The original appears to be a Klaxon, possibly from a Model T. A replacement was located which, though smaller and newer, closely resembles the original horn. It is also a 6 volt unit matching the M-3 electrical system. Unlike the original, it has a mechanical linkage that sounds the horn. In order to trigger the horn a cable linkage had to be run to the horn, and to avoid the front wheels, the location of the horn had to be moved from the left to the right fender.
The rear tail lights appear to have been from a 1918-24 Model T Ford. One of the tail lights was still in place while the other one was missing. A nearly identical Model T tail light was located for the unit. It varied in that it had a side light to illuminate a license plate and it had a single contact socket rather than a dual contact socket. These two features were changed to match the existing socket.
As in most earlier cars, the M-3 has a DC generator rather than an alternator producing AC current which is then rectified to DC. The alternator is used today because it has no segmented commutator to maintain, using solid slip rings instead. However, generator has the ability to adjust the position of the contact brushes, thus changing the power output. The alternator has a varying voltage output based only on it's rotational speed. Thus, the generator on different cars had different brush settings based on how many appliances the operator normally used. The generator on the M-3 was set to provide enough sustained power to operate the engine and lights at all times for safety.
The original dashboard voltage meter was repaired and reinstalled.
After being moved to eastern Pennsylvania, total disassembly of the car began. Roof, seats, walls, doors, axles, spring rigging, controls, floors and anything else that could be removed was removed. The unit was stripped down to the frame which was completely sandblasted, primed and painted.
The steel composing the frame and substructure of the body was retained as no replacement was needed. Some broken rivets were replaced by either welding or bolting as appropriate.
During the disassembly most wood trim was damaged too much to be reused. This was due to the fact that the screws that held trim on were puttied over and virtually impossible to locate. Many of the side and roof boards were cracked, broken or warped. This led to the decision to replace the body wood with new material. With the replacements the thickness of the boards was increased from 3/8 to 7/16 for additional strength. All new wood was oak. The seat boards were similarly replaced with new oak.
The floor inside the car was covered with a strange variety of boards, some tongue and groove and others not, some pine and some oak. The boards extended only as far as the seats and not to the outside wall. The boards were in poor condition with numerous holes, cracks, breaks and rot. The flooring was covered with a layer of conveyor belt (5 layers of canvas with rubber coating.) This left only the conveyor belt on the floor areas under the seats. A complete replacement of the floor wood was unavoidable. The new wood was oak and extended to the walls of the passenger compartment. They were covered with an epoxy coating to extend the life of the wood.
Another peculiarity of the car is that many boards were held in place by 1/4-18 (18 threads per inch) bolts with square nuts. As 1/4-18 is effectively impossible to find today, the bolts were replaced by modern 1/4-20 (20 threads per inch) bolts.
The canvas sides of the car were torn in several places and patched back together. It was decided to replace the canvas. As the original canvas could be torn by a simple kick or stray tool, the new canvas was reinforced by a backing of polymer FRP board. The sides of the car were primed and painted with four coats of gloss red.
On the roof, the replacement canvas was used alone as was the case originally. During installation the canvas was stretched and stapled into place. it was then coated with boiled linseed oil to fill in the pores in the canvas. This prevents the paint from soaking into the canvas and cracking as the canvas flexed. The roof was painted it original silver color.
The original lettering on the side of the car was used as a pattern to create stencils that exactly recreated the original gold lettering.
Another item the car was missing was the engineer's seat. The only remnant of the seat were four bolt holes on the floor in the appropriate locations. Unfortunately, no photos of the interior of the M-3 showing the original seat have been found. The restoration crew created an engineer's seat that is a "guestimate" of what the original may have looked like. It uses the adjustable top of a piano type stool and a steel base bolted in the original's location.
One element long missing from the car, even in its common carrier days, was the pilot (cowcatcher) sported on the front of the car in its earliest photo. The restorers elected to recreate this unique element based on historic photos. The new pilot sets a couple inches above the original to avoid any complications from switches, crossing and other obstructions that may have brought an end to the original. As noted above the pilot included a recreation of the original crank starter. The pilot is removable should the car need to be shown in its later configuration.
The side glass assemblies in the car are original, both the wood and glass. They likely date to the 1924 construction of the car as the glass has the characteristic flowing nature of glass of that period. The wood components were sanded, filled, primed and painted with four coats.
Several photos of the car show that it featured a Model-T style split front windshield. This means the windshield frame was hinged and the top pane of glass could be rotated down behind the bottom pane, to allow ventilation. Sometime after 1956 the windshield was replaced with a single solid pane. The recesses for the side hinges remained. During the restoration, the split windshield was also returned to the car. Reproduction Model T hinges were acquired and a new windshield frame built around them. The reproduction had to be cut, lengthened and rewelded to fit the M-3. After the frame was built it was found that both hinges were right hinges, but this caused no functional problems and only added to the quirkiness the railcar already had.
The front of the passenger compartment was made of a fairly light weight sheet steel. The sheet was actually in two sections. In the process of removal the sheets were bent and twisted. It was decided to replace them with a single sheet of steel, much thicker for safety and durability. Whereas the original steel could be bent by impact with a passenger or tool, the new one will remain rigid and straight. A rear view mirror was also added for safety
The old hood was in bad shape. The rear (firewall) end of the hood did not rest on anything. However the steel parts of the hood were original Maxwell materials and so needed to be salvaged as much as possible. A wood frame was built on the firewall to match the hood shape and give the rear of the hood something to sit on. The hood was reconstructed with new sheet metal, but with the original louvers welded into place. Patterns for the hood were first constructed in cardboard and then the cardboard was used as a pattern to construct the metal hood pieces.
New front fenders were made from stainless steel to give them additional service life. The firewall, which is made in the original Maxwell shape was redone using new varnished oak.